13 Current model for spindle positioning requires attachment of the microtubule (MT) motor 14 cytoplasmic dynein to the cell cortex, where it generates pulling force on astral MTs to effect 15 spindle displacement. How dynein is anchored by cortical attachment machinery to generate 16 large spindle-pulling forces remains unclear. Here, we show that cortical clustering of Num1, the 17 yeast dynein attachment molecule, is limited by Mdm36. Overexpression of Mdm36 results in an 18 overall enhancement of Num1 clustering but reveals a population of dim Num1 clusters that 19 mediate dynein-anchoring at the cell cortex. Direct imaging shows that bud-localized, dim Num1 20 clusters containing only ~6 copies of Num1 molecules mediate dynein-dependent spindle 21 pulling via lateral MT sliding mechanism. Mutations affecting Num1 clustering interfere with 22 mitochondrial tethering but not dynein-based spindle-pulling function of Num1. We propose that 23 formation of small ensembles of attachment molecules is sufficient for dynein anchorage and 24 cortical generation of large spindle-pulling force. 25 26 4 al. , 2010; Lackner et al., 2013). Although it is well-accepted that dynein exerts spindle pulling 53 force at cortical Num1 sites, the abundance and heterogeneity of Num1 patches along the cell 54 cortex (Heil-Chapdelaine et al., 2000; Omer et al., 2018; Schmit et al., 2018) has made it 55 impossible to follow the effects of astral MT plus end interaction with individual cortical Num1 56 sites, a prerequisite for understanding how clustering might impact dynein force amplification. 57 To our knowledge, contacts between astral MT plus end and individual Num1 foci have not 58 been observed for MT sliding, the in vivo hallmark of dynein-mediated spindle pulling (Adames 59 and Cooper, 2000; Yeh et al., 2000), hence the size of Num1 clusters required for this classic 60 dynein-dependent microtubule-cortex interaction remains unknown. Additionally, recent work61 shows that organelles such as mitochondria and endoplasmic reticulum (ER) are involved in 62 regulating Num1 cluster formation: a subset of cortical Num1 clusters appears to require 63 mitochondria for their assembly (Kraft and Lackner, 2017), whereas another population requires 64 the ER tethering proteins Scs2/Scs22 for their distribution throughout the cell cortex (Chao et 65 al., 2014; Omer et al., 2018). The general hypothesis emerging from these studies is that 66 distinct populations of Num1 clusters might exist at the cell periphery, but whether different 67 pools of Num1 could be performing different Num1 functions -namely, dynein anchoring and 68 mitochondrial tethering -remains a total mystery.69 Here, we set out to characterize the role of Mdm36 in Num1 clustering and found that, in 70 contrast to the prevailing notion for dynein-anchoring proteins, enhancing Num1 clustering 71 unexpectedly reduces dynein recruitment to the cell cortex, but without affecting dynein function 72 in spindle positioning. We report direct observation of MT sliding occurring upon...